Rename columns using `starts_with()` where new pre

2020-02-06 07:44发布

问题:

In R, I want to rename all the columns that starts with some prefix (say "oldprefix1", "oldprefix2", "oldprefix3", ... to "newprefix1", "newprefix2", "newprefix3", ...) inside a function. The following code works:

change = function(df) {
    select(df, newprefix = starts_with('oldprefix') )
}
change(test)

But, I would like to pass a string with the new prefix as parameter to the function:

change2 = function(df, prefix) {
    dots = paste0(prefix," = starts_with('oldprefix')"
    select_(df, dots)
}
change2(test, "newprefix")

I have tried using select_() and .dots, but I cannot get it to work together with the starts_with() function. I get the error Error in eval(expr, envir, enclos) : could not find function "starts_with".

回答1:

The option would be to use rename_at

mtcars %>% 
    rename_at(vars(starts_with('m')), funs(paste0('prefix', .)))

For changing an old name, use sub

change2 <- function(df, oldpref, newpref) {
  df %>%
       rename_at(vars(starts_with(oldpref)), funs(sub(oldpref, newpref, .)))


 }

change2(mtcars, "m", "newprefix") %>%
       names
#[1] "newprefixpg" "cyl"         "disp"        "hp"          "drat" 
#[6]   "wt"          "qsec"        "vs"          "am"          "gear" 
#[11] "carb"    


标签: r dplyr rename